Predicting Alzheimer’s Disease Using Deep Neuro-Functional Networks with Resting-State fMRI

Resting-state functional connectivity has been widely used for the past few years to forecast Alzheimer’s disease (AD). However, conventional correlation calculation does not consider different frequency band features that may hold brain atrophies’ original relationships. Previous works focuses on low-order neurodynamics and precisely manipulates mono-band span of resting-state magnetic imaging (rs-fMRI). They specifically use rs-fMRI, leaving out high-order neurodynamics. By creating a neuro-dynamic network employing several levels rs-fMRI time-series data, such as slow4, slow5, full-band ranges (0.027 0.08 Hz), (0.01 0.027 we suggest an automated AD diagnosis system address these challenges. It combines multiple customized deep learning models provide unbiased evaluation, tenfold cross-validation is observed We have determined differentiate disorders from NC, entire slow4 referred higher lower approaches, are applied. The first method uses SVM KNN deal with diseases. second Alexnet Inception blocks datasets ADNI organizations. also tested other machine approaches by modifying various parameters attained good accuracy levels. Our proposed model achieves performance using three bands without any external feature selection. results show our (96.61%)/AUC (0.9663) achieved in differentiating subjects normal controls. Furthermore, accuracies classifying stages potentiality clinical value prediction.